Forecasting Energy Utilization in Residential Areas through Machine Learning Techniques

Abstract

Analyzing and forecasting energy use in residential areas and buildings is critical for promoting sustainability and optimising energy usage. This paper describes a unique way for exact energy consumption analysis and forecasting that employs machine learning techniques such as linear regression and ridge regression. The proposed model intends to provide important insights and forecasts by combining historical energy data with relevant parameters such as building attributes, weather conditions, and tenant behaviour, improving energy usage in residential settings. The expected prototype is tested using a dataset accessible on Kaggle. The findings provided in the study show that our suggested model surpasses the state-of-the-art system, with accuracy rates of 66.56 percent for linear regression and 70.12 percent for ridge regression, respectively.